Abstract
Lysosomal dysfunction plays a central role in the pathogenesis of several neurodegenerative disorders, including Parkinson's disease (PD). Several genes linked to genetic forms of PD, including leucine-rich repeat kinase 2 (LRRK2), functionally converge on the lysosomal system. While mutations in LRRK2 are commonly associated with autosomal-dominant PD, the physiological and pathological functions of this kinase remain poorly understood. Here, we demonstrate that LRRK2 regulates lysosome size, number and function in astrocytes, which endogenously express high levels of LRRK2. Expression of LRRK2 G2019S, the most common pathological mutation, produces enlarged lysosomes and diminishes the lysosomal capacity of these cells. Enlarged lysosomes appears to be a common phenotype associated with pathogenic LRRK2 mutations, as we also observed this effect in cells expressing other LRRK2 mutations; R1441C or Y1699C. The lysosomal defects associated with these mutations are dependent on both the catalytic activity of the kinase and autophosphorylation of LRRK2 at serine 1292. Further, we demonstrate that blocking LRRK2’s kinase activity, with the potent and selective inhibitor PF-06447475, rescues the observed defects in lysosomal morphology and function. The present study also establishes that G2019S mutation leads to a reduction in lysosomal pH and increased expression of the lysosomal ATPase ATP13A2, a gene linked to a parkinsonian syndrome (Kufor–Rakeb syndrome), in brain samples from mouse and human LRRK2 G2019S carriers. Together, these results demonstrate that PD-associated LRRK2 mutations perturb lysosome function in a kinase-dependent manner, highlighting the therapeutic promise of LRRK2 kinase inhibitors in the treatment of PD.
Introduction
The lysosome plays a central role for the living cell, acting as its primary degradative compartment. Serving as the final point of convergence of several proteolytic pathways, it digests extracellular material that has been internalized through endocytosis and recycles intracellular components through autophagy (1–4). Additionally, lysosomes coordinate diverse physiological processes including maintenance of the plasma membrane, lipid homeostasis, pathogen defense, cell signaling and apoptosis (5–7). As such, lysosomal defects can have widespread consequences for cell health and survival, leading to the accumulation of cellular debris and defective metabolism and, ultimately, to cell death. The consequences of such dysfunction are highlighted by the many diseases characterized by the accumulation of indigestible materials, including neurodegenerative disorders such as Parkinson's disease (PD) (8–10).
Parkinson's disease results from the deterioration of the nigrostriatal dopaminergic system, which undergoes loss of dopaminergic neurons and increased gliosis within the substantia nigra (11). Another hallmark of PD is the accumulation of Lewy bodies, which are proteinaceous cytoplasmic inclusions rich in α-synuclein, suggesting defective protein clearance may contribute to pathogenesis (12). In support of this, post-mortem analyses of PD patient brain samples report reductions in lysosomal markers and increased accumulation of autophagosomes (13–16). Emerging genetic data further strengthen the link between lysosomal dysfunction and PD, as several key genes linked to familial forms of PD and inherited parkinsonian syndromes converge on the lysosomal pathway, including SNCA, GBA, ATP13A2, VPS35 and leucine-rich repeat kinase 2 (LRRK2) (8,17).
Mutations within LRRK2 are a common known cause of autosomal-dominant PD, accounting for ∼1% of sporadic and 5% of familial cases (18). LRRK2 encodes a large multi-domain protein containing a GTPase domain, kinase domain and several potential protein–protein interaction domains (19,20). The majority of its identified pathogenic mutations are located within the central catalytic domains, including the most common mutation associated with LRRK2 (G2019S) (21–23). Increased LRRK2 kinase activity has been proposed to contribute to pathogenesis, suggesting the therapeutic potential of LRRK2 kinase inhibitors in the treatment of PD (24,25). The physiological and pathological functions of LRRK2, however, remain unclear. LRRK2 has been reported to localize, at least to some extent, to vesicular structures such as endosomes and lysosomes (26,27) and several reports suggest LRRK2 may function as a regulator of the endolysosomal system (26,28–34). However, a precise understanding of how LRRK2 modulates the lysosomal pathway and of how its PD-associated mutations perturb this function is lacking.
Here, we demonstrate LRRK2 regulates lysosome size, number and function in primary astrocytes, a cell type which endogenously expresses high levels of LRRK2. Our results show that expression of PD-associated LRRK2 variants results in enlarged lysosomes, a phenotype that can be prevented genetically by elimination of kinase activity (D1994A co-mutation) or LRRK2 autophosphorylation (S1292A co-mutation). We demonstrate that expression of the G2019S mutation inhibits lysosomal degradation of long-lived proteins and reduces lysosomal pH. Additionally, we establish that inhibiting LRRK2’s kinase activity, using the potent and selective inhibitor PF-06447475, rescues defects in lysosome morphology and lysosomal degradation seen with the pathogenic G2019S LRRK2 mutation. Finally, our results illustrate that levels of the lysosomal ATPase, ATP13A2, are increased in LRRK2 G2019S brain samples from mice and from human LRRK2 G2019S carriers, suggesting a potential link between two PD-associated proteins. Together, our observations identify a conserved phenotype seen with pathogenic LRRK2 mutations in the lysosomal pathway, establish that these lysosomal perturbations depend on LRRK2’s kinase activity and on autophosphorylation of LRRK2 at S1292 and provide a mechanism by which these lysosomal defects can be corrected. Further, the present work provides increased support for the therapeutic potential of LRRK2 kinase inhibitors in the treatment of PD.
Results
LRRK2 G2019S mutation regulates lysosome size and morphology
We chose primary mouse astrocytes as our main model to assess LRRK2 function given their high endogenous LRRK2 expression compared with other cell types assessed [Fig. 1A and (35)]. To explore the role of LRRK2 in the endocytic pathway, we utilized astrocytes from non-transgenic mice, transgenic mice overexpressing the wild-type mouse LRRK2 protein (Wt-LRRK2 Tg), transgenic mice overexpressing the mutant form of mouse LRRK2 (LRRK2 G2019S Tg) associated with autosomal-dominant PD and LRRK2 KO mice. We began by looking for any gross differences in early endosome and lysosome morphology with LRRK2 overexpression, mutation or deletion. While immunofluorescence staining of early endosomes (EEA1) showed no overt differences in morphology across genotypes, lysosomal staining (using the marker LAMP2) revealed enlarged structures with a tight, perinuclear distribution specifically in LRRK2 G2019S Tg astrocytes (Fig. 1B). Live-cell imaging experiments using LysoTracker Red corroborated this effect (Fig. 1C). To quantify changes in lysosome size, number and localization, astrocytes were transduced with CellLight-Lysosome-GFP BacMam virus to allow selective characterization of lysosomes (apart from other acidic compartments such as early/late endosomes) and to prevent possible reductions in lysosomal volume induced by fixation. Lysosomes in LRRK2 G2019S Tg astrocytes were significantly larger in size, with an increase of nearly 2-fold, on average, compared with those from non-transgenic or LRRK2 KO animals and compared with those overexpressing LRRK2 at similar levels, illustrating that this effect on lysosome size was due to LRRK2 G2019S mutation rather than overexpression itself (Fig. 1D and Supplementary Material, Fig. S2A). Lysosomes from LRRK2 Tg astrocytes (G2019S and wild-type) also localized closer to the nucleus compared with the non-transgenic astrocytes (Supplementary Material, Fig. S1A–C). Further, stark differences in lysosome number were observed across the genotypes (Fig. 1D). The number of lysosomes per cell was significantly reduced in Wt-LRRK2 Tg astrocytes compared with non-transgenic astrocytes, which was further diminished upon LRRK2 G2019S mutation. Conversely, the number of lysosomes per cell approximately doubled in LRRK2 KO astrocytes compared with those from non-transgenic animals.
LRRK2 regulates the size and number of lysosomes in primary mouse astrocytes. (A) LRRK2 protein levels were compared across different cell types including HEK 293 cells, primary mouse astrocytes from non-transgenic and LRRK2 KO animals, primary mouse microglia from non-transgenic animals and primary mouse cortical neurons (DIV 14). Shown is a representative immunoblot of LRRK2 with actin loading control. Fluorescence signals of immunoblots from multiple experiments were quantified and the LRRK2 signal was normalized to actin and expressed as a % compared with HEK 293 cells; n = 4. (B) Astrocytes from non-transgenic, Wt-LRRK2 Tg, LRRK2 G2019S Tg and LRRK2 KO mice were fixed and stained with antibodies against EEA1 and Lamp2. (C) Endolysosomal morphology was assessed in these cells using LysoTracker Red. (D) Astrocytes from non-transgenic, Wt-LRRK2 Tg, LRRK2 G2019S Tg and LRRK2 KO mice were transduced with CellLight-Lysosome-GFP baculovirus and imaged live. Lysosomal size and number were quantified using ImageJ; Nontg (140 cells), Wt-LRRK2 Tg (86 cells), LRRK2 G2019S (91 cells), non-transgenic cells imaged with LRRK2 KO (Nontg) (101 cells) and LRRK2 KO (103 cells) were analyzed in three independent experiments (n = 3); scale bar = 5 µm. LRRK2 KO astrocytes were not imaged during the same experiments as the Wt-LRRK2 Tg and LRRK2 G2019S Tg cells (due to different timings of our primary cultures across the different genotypes) but were imaged at the same time with a separate set of non-transgenic astrocytes, therefore we show these data as two separate bars on the chart. Error bars correspond to SEM and P-values: Student's t-test or one-way ANOVA, Bonferroni multiple comparison test; *P < 0.05; **P < 0.01; ***P < 0.001, ****P < 0.0001.
These changes in lysosome number were mirrored by changes in the levels of lysosomal membrane proteins Lamp1 and Lamp2 in these cells (Fig. 2A). Further, we measured Lamp2 levels across genotypes in the mouse midbrain to determine whether these changes were also detected from brain samples. Lamp2 protein levels were significantly increased in LRRK2 KO mouse brain samples and showed a trend toward reduction in LRRK2 G2019S Tg midbrain samples, consistent with lysosomal alterations seen in culture (Fig. 2B). To determine whether these changes were also seen in human patients that carried the LRRK2 G2019S mutation, we performed western blot analysis of prefrontal cortex samples from five control, five sporadic PD and five LRRK2 G2019S mutation carriers. While there was no statistically significant change, analysis of sporadic PD and LRRK2 G2019S patient samples showed a trend toward reduced Lamp2 levels compared with control patient samples (Fig. 2C). Together, these data indicate that LRRK2 G2019S mutation alters lysosome morphology, leading to the formation of enlarged lysosomes clustered near the nucleus. Further, these results suggest that LRRK2 expression and mutation modulate the lysosomal capacity of the cell.
LRRK2 regulates the levels of lysosomal proteins in primary mouse astrocytes, mouse midbrain and human post-mortem samples. (A) Lamp1 and Lamp2 protein levels were compared across non-transgenic, Wt-LRRK2 Tg, LRRK2 G2019S Tg and LRRK2 KO primary mouse astrocytes. Shown are representative immunoblots also depicting LRRK2 protein levels across samples and using actin as a loading control. Fluorescence signals of immunoblots from multiple experiments were measured and the Lamp1 and Lamp2 signals were normalized to actin and expressed as a % of that seen in non-transgenic astrocytes; n ≥ 3. (B) Lamp2 protein levels were measured from mouse midbrain samples across genotypes, actin normalized and expressed as a % of that detected in non-transgenic midbrain samples; n = 3 animals per genotype. (C) Lamp2 levels were measured in age-matched, human post-mortem samples from the prefontal cortex of control, LRRK2 G2019S carriers and idiopathic PD patients. Lamp2 signal was normalized to actin and expressed as a % of that detected in control samples; n = 5 patients per condition. Error bars correspond to SEM and P-values: Student's t-test or one-way ANOVA, Bonferroni multiple comparison test; *P < 0.05; **P < 0.01; ***P < 0.001.
LRRK2 G2019S mutation impairs lysosome function and lowers endolysosomal pH
To determine whether there were any functional consequences of the enlarged lysosome phenotype seen with G2019S mutation, we employed multiple approaches to assess lysosome functionality with LRRK2 overexpression, G2019S mutation or LRRK2 deletion. Previous studies have examined the role of G2019S mutation and LRRK2 deletion in autophagic degradation and have yielded conflicting results as to whether loss of LRRK2 or expression of its pathogenic mutations enhance or inhibit autophagy (28,32,36–42). We also saw variable effects of LRRK2 manipulations on autophagy in our cellular models, as assessed by LC3-II and p62 levels across astrocyte genotypes (data not shown), and thus chose to focus on whether LRRK2 regulated lysosomal function in a more general manner. First, we examined the proteolytic capacity of these cells by measuring the rate of intracellular protein degradation. We found that the total degradation of long-lived proteins under basal conditions was very similar across genotypes in our astrocyte cultures (Fig. 3A). However, when we determined the percentage of protein degradation occurring within lysosomes (the component that is sensitive to inhibition by chloroquine), we found that LRRK2 G2019S Tg astrocytes alone had significantly decreased lysosomal activity (Fig. 3B). It should be noted that while not significant, there was a slight reduction in lysosomal activity in astrocytes from Wt-LRRK2 Tg astrocytes, which could suggest that overexpression of LRRK2 itself reduces lysosomal function, albeit to a lesser extent that G2019S mutation. These results suggest an impairment of lysosomal function upon LRRK2 G2019S mutation, an effect which may be compensated for by additional, non-lysosomal modes of degradation.
Lysosomal function is perturbed by LRRK2 manipulation. (A and B) Primary astrocytes from non-transgenic, Wt-LRRK2 Tg, LRRK2 G2019S Tg and LRRK2 KO mice were labeled with [3H]-valine for 24 h. After three rounds of washing, cells were treated for 24 h with neurobasal medium containing unlabeled valine (A) and supplemented with 200 nM chloroquine (B) to measure lysosomal degradation. Protein degradation was measured as the amount of acid-precipitatable radioactivity released into the media and is expressed as the percentage of radioactivity released in the media compared with that remaining in the cell lysate; n ≥ 3. (C) Cathepsin B activity was assessed across genotypes using a fluorescence-based cathepsin B detection kit. Cathepsin B activity was measured and is expressed as a % of that seen in non-transgenic astrocytes; n = 3. (D and E) Astrocytes were labeled with 2 µM LysoSensor Yellow/Blue DND-160 for 20 min, rinsed 3 times with isotonic solution and incubated with pH buffers to generate a standard curve (D) or with isotonic solution to measure basal endolysosomal pH across genotypes (E). After 10 min, fluorescence was measured and endolysosomal pH was quantified as the ratio of light excited at 340 nm over 380 nm, detected at a 527 nm emission; n ≥ 3. Error bars correspond to SEM and P-values: One-way ANOVA, Bonferroni multiple comparison test; *P < 0.05; **P < 0.01; ***P < 0.001, ****P < 0.0001.
We next examined cathepsin B activity to identify potential mechanistic explanations for the impaired lysosomal activity seen with LRRK2 G2019S mutation. Cathepsin B activity was measured in intact, living cells and was markedly reduced specifically in astrocytes expressing the LRRK2 G2019S mutation (Fig. 3C). As defects in lysosomal function and specifically, cathepsin B activity, were seen in LRRK2 G2019S Tg astrocytes, we postulated that changes in lysosomal pH could account for compromised lysosomal enzyme function, as lysosomal protease activity is extremely sensitive to pH changes. We measured lysosomal pH using a ratiometric lysosomal pH indicator dye (LysoSensor Yellow/Blue DND-160), which exhibits pH-dependent changes in excitation and emission. The ratio of light excited at 340/380 nm was converted to specific pH values by running parallel calibration curves in cells whose lysosomal pH was adjusted to different values from pH 4.0 to 6.0 (Fig. 3D). Measurement of average lysosomal pH across astrocyte genotypes demonstrated that lysosomal pH was intriguingly reduced selectively in LRRK2 G2019S-expressing cells (Fig. 3E). This reduction in lysosomal pH is in contrast to elevations in lysosomal pH that have been previously reported in cellular models of Alzheimer's disease (43–45) and could be informative in terms of the unique consequence of LRRK2 G2019S mutation on lysosome function.
LRRK2 localizes to lysosomes and regulates the size of lysosomes through its kinase activity
As LRRK2 G2019S mutation perturbed lysosome size and function, we next wanted to determine where LRRK2 localized in primary astrocytes. Previous studies have not lead to a consensus about its subcellular localization, with LRRK2 reported to localize to the cytoplasm, Golgi apparatus, mitochondria, plasma membrane, synaptic vesicles, early/late endosomes and lysosomes (27,40,46–48). We transduced non-transgenic astrocytes with a lentivirus encoding GFP-tagged human LRRK2 (Leuven Viral Vector Core) to assess LRRK2 localization in cultured astrocytes (Fig. 4A). GFP-tagged LRRK2 showed low colocalization with the early endosomal marker Rab5 [colocalization coefficient (R) = 0.134] and increased colocalization with the late endosomal marker Rab7 (R = 0.349) (Fig. 4B). However, GFP-tagged LRRK2 colocalized the strongest with the lysosomal marker Lamp2 (R = 0.857). To ensure this localization was not an artifact of overexpression, colocalization of LRRK2 with lysosomal markers was also assessed using antibodies against endogenous LRRK2 and against the FLAG epitope present on the LRRK2 transgene in the transgenic mice (Supplementary Material, Fig. S3A and B). These immunofluorescence studies also showed strong colocalization of LRRK2 with Lamp2, suggesting that in cultured astrocytes, LRRK2 localizes primarily to lysosomes.
LRRK2 localizes to lysosomes in astrocytes and its kinase activity regulates the size of lysosomes. (A) Astrocytes from non-transgenic mice were transduced with a lentivirus encoding GFP-LRRK2, fixed 3 days after transduction and stained for various endocytic markers. Shown are representative images; scale bar = 5 µm (B) The percentage of GFP-LRRK2 positive structures colocalizing with different endosomal proteins was quantified. (C) Astrocytes from wild-type mice were transduced with a lentivirus encoding GFP alone, GFP-LRRK2, GFP-LRRK2 G2019S or GFP-LRRK2 G2019S D1994A, and images were taken over 7 days. Lysosomes were visualized with LysoTracker Red and cells were imaged live using a Spinning-Disk confocal microscope. The expression level of the various constructs used is quantified in (D). The amount of cells with lysosomes with a diameter greater than 3 µm was quantified using ImageJ, and a representative time course is shown (E); n ≥ 21 cells per time point, per lentiviral construct expressed.
To gain more insight into how LRRK2 G2019S mutation leads to enlarged lysosomes, we transduced non-transgenic astrocytes with GFP alone, GFP-tagged wild-type LRRK2, GFP-LRRK2 G2019S and GFP-LRRK2 G2019S-D1994A (which eliminates kinase activity) and followed changes in lysosome size over time (Fig. 4C). We verified equal expression of LRRK2 constructs by quantifying GFP fluorescence across cells examined (Fig. 4D). Exogenous expression of GFP-LRRK2 G2019S lead to the formation of enlarged lysosomes, an effect that was observed 3 days post viral transduction which correlated with when lentiviral expression reached its peak (Fig. 4E). This effect was not observed with expression of wild-type LRRK2 nor with the kinase-dead version of LRRK2 G2019S, suggesting that the formation of enlarged lysosomes requires LRRK2’s kinase activity itself.
Selective inhibition of LRRK2’s kinase activity rescues defects in lysosome morphology and function
As genetic elimination of LRRK2’s kinase activity (G2019S D1994A co-mutation) prevented the formation of enlarged lysosomes, we hypothesized that pharmacological inhibition of its kinase activity may rescue the defects in lysosome size and functionality seen with G2019S mutation. We first verified that treatment with the selective and potent LRRK2 kinase inhibitor, PF-06447475 (49), inhibited phosphorylation of LRRK2 at Ser935 and Ser1292 and did not alter total LRRK2 protein levels in our primary astrocyte cultures (Supplementary Material, Fig. S4C and D). Next, we transduced astrocytes with GFP-LRRK2 G2019S lentivirus, treated cells with vehicle alone or with a selective and potent LRRK2 kinase inhibitor, PF-06447475 (49), and visualized their lysosomes using LysoTracker Red 3 days later. Vehicle-treated cells exhibited the characteristic enlarged lysosomes seen upon LRRK2 G2019S mutation (Fig. 5A). In contrast, cells treated with the LRRK2 kinase inhibitor had a marked reduction in lysosome size and an increase in lysosome number. GFP-LRRK2 G2019S itself also redistributed to these smaller lysosomes. To ensure these effects were on target, we performed dose-response analysis and calculated an IC50 = 3.4 nM ± 1.6 nM (Fig. 5B), a value that is well-aligned with more direct pharmacological endpoints (49). Additionally, inhibition of LRRK2’s kinase activity significantly increased lysosomal degradation in Wt-LRRK2 Tg and G2019S Tg astrocytes (Fig. 5C), suggesting that LRRK2 inhibitor treatment can improve lysosome functionality. Together, these data suggest LRRK2’s kinase activity itself modulates lysosome size and lysosome function and highlight the therapeutic potential of LRRK2 kinase inhibitors in correcting lysosomal defects seen with LRRK2 mutation.
Selective inhibition of LRRK2’s kinase activity rescues defects in lysosome morphology and function. (A and B) Astrocytes from non-transgenic mice were transduced with a lentivirus encoding GFP-LRRK2 G2019S and treated with varying concentrations of vehicle or PF-064474475. Lysosomes were visualized with LysoTracker Red, and cells were imaged using a Spinning-Disk confocal microscope. Shown are representative images after 3 days treatment with vehicle or 500 nM PF-064474475 (A); scale bar = 5 µm. (B) Dose-response curves for rescue of the enlarged lysosome phenotype were generated and an IC50 value of 3.4 nM ± 1.6 nM was determined using GraphPad Prism software; Vehicle n = 176 cells, 0.1 nM PF-‘475 n = 105 cells, 1 nM PF-‘475 n = 100 cells, 10 nM PF-‘475 n = 111 cells, 100 nM PF-‘475 n = 115 cells, 500 nM PF-‘475 n = 120 cells, 3 independent experiments. (C) Primary astrocytes from non-transgenic, Wt-LRRK2 Tg, LRRK2 G2019S Tg and LRRK2 KO mice were treated with vehicle or 500 nM PF-064474475 for 24 h and labeled with [3H]-valine in media containing vehicle or 500 nM PF-064474475 for an additional 24 h. After three rounds of washing, cells were treated for 24 h with neurobasal medium containing unlabeled valine, vehicle or 500 nM PF-064474475 and supplemented with water or 200 nM chloroquine to measure lysosomal degradation. Protein degradation was measured as the amount of acid-precipitatable radioactivity released into the media and is expressed as the percentage of radioactivity released in the media compared with that remaining in the cell lysate; n ≥ 3. Error bars correspond to SEM and P-values: Student's t-test or one-way ANOVA, Bonferroni multiple comparison test; *P < 0.05 and **P < 0.01.
Phosphorylation of LRRK2 itself promotes the enlarged lysosome phenotype seen with its pathogenic mutations
Our studies thus far have focused on the G2019S mutation, as it is the most common LRRK2 mutation. We next wanted to explore how LRRK2’s kinase activity regulates the lysosomal pathway and whether this was a more common phenotype associated with other pathogenic LRRK2 mutations. While its in vivo substrates remain unknown, an autophosphorylation site within LRRK2 (S1292) was recently identified whose phosphorylation correlated well with LRRK2’s kinase activity and was increased by PD-associated LRRK2 mutations (50,51). To gain mechanistic insight into how LRRK2 modulates the lysosomal pathway, we assessed whether other PD-associated mutations within LRRK2 also lead to enlarged lysosomes and whether phosphorylation at S1292 could regulate this effect. To do this, we generated a series of GFP-tagged LRRK2 constructs carrying pathogenic mutations and additional mutations that rendered LRRK2 kinase-dead (D1994A) or prevented S1292 phosphorylation (S1292A). We verified that PD-associated mutations within LRRK2 (G2019S, R1441C and Y1699C) increased LRRK2 S1292 phosphorylation by western blot analysis (Supplementary Material, Fig. S4A and B). Further, we examined their expression and localization in HEK293 cells by immunofluorescence and observed that, unlike in primary astrocyte cultures, GFP-tagged LRRK2 localized to the cytoplasm (Fig. 6A), suggesting that LRRK2’s localization may vary according to cell-type. When these mutants were expressed in primary astrocytes, we found that expression of LRRK2 R1441C and Y1699C mutants also increased the size of lysosomes (Fig. 6B). It is interesting to note, however, that while enlarged lysosomes were observed with these additional pathogenic mutations, expression of these mutants did not have as profound an effect on lysosome size and number as the G2019S mutation (Supplementary Material, Fig. S2B and C). The enlarged lysosome phenotype was completely reversed by D1994A co-mutation, suggesting that like G2019S, these additional PD-associated mutations also regulate lysosome size in a kinase-dependent manner (Fig. 6B and C). Further, preventing autophosphorylation at S1292 significantly reduced the percentage of cells with enlarged lysosomes, albeit to a lesser extent than the kinase-dead mutation. Together, these results suggest that changes in lysosome size are a more generalized effect seen with PD-associated LRRK2 mutations and that phosphorylation of LRRK2 itself modulates lysosome morphology.
Phosphorylation of LRRK2 itself contributes to the enlarged lysosome phenotype seen with its pathogenic mutations. (A) HEK 293 cells were transfected with GFP alone or various GFP-tagged mutant LRRK2 constructs and imaged live. Representative examples of each condition are shown. (B) Non-transgenic astrocytes were transfected with the same GFP-LRRK2 mutant constructs, lysosomes were visualized with LysoTracker Red, and imaged live. Shown are representative examples of each condition. Expression level of GFP-tagged constructs was quantified (C) and the amount of cells with lysosomes with a diameter greater than 3 µm was quantified under these different conditions in ImageJ (D); GFP (72 cells), LRRK2 (91 cells), LRRK2 S1292A (72 cells), LRRK2 G2019S (80 cells), LRRK2 G2019S S1292A (91 cells), LRRK2 G2019S D1994A (103 cells), LRRK2 R1441C (90 cells), LRRK2 R1441C S1292A (94 cells), LRRK2 R1441C D1994A (129 cells), LRRK2 Y1699C (84 cells), LRRK2 Y1699C S1292A (91 cells), LRRK2 Y1699C D1994A (82 cells), n = 3 independent experiments. Scale bar = 5 µm; error bars correspond to SEM and P-values: One-way ANOVA, Bonferroni multiple comparison test; *P < 0.05; **P < 0.01; ***P < 0.001, ****P < 0.0001.
ATP13A2, a lysosomal p-type ATPase, is upregulated in LRRK2 G2019S mouse and human brain samples
We next wanted to gain insight into how LRRK2 mutation modulated the lysosomal pathway by examining whether LRRK2 regulated additional key lysosomal components. In particular, the reduction in lysosomal pH seen with LRRK2 G2019S mutation was especially intriguing and could provide unique mechanistic insight. To this end, we focused on lysosomal proteins reported to play a role in maintaining proper lysosome pH including lysosomal V- and P-type ATPases and ion channels. Of the proteins assessed, western blot analysis revealed upregulation of ATP13A2 in LRRK2 G2019S Tg primary astrocyte cultures and mouse midbrain lysates (Fig. 7A and B). ATP13A2 levels were also increased in midbrains from LRRK2 KO animals, but this effect was lost when corrected for the increase in lysosomes and lysosome-associated membrane proteins seen in these animals (compared with Fig. 2B). To confirm the specificity of this antibody against ATP13A2, we performed siRNA knockdown of ATP13A2 using three different siRNA oligos and found that all of them reduced the signal detected by our ATP13A2 antibody to varying extents (Supplementary Material, Fig. S5). ATP13A2 is a lysosomal P-type ATPase that has been reported to modulate lysosomal pH and to be necessary for proper lysosome function. A link between LRRK2 and ATP13A2 is especially tantalizing, as ATP13A2 has also been associated with young-onset forms of PD and with Kufor–Rakeb syndrome, a parkinsonian syndrome (52). Previous studies assessing ATP13A2 expression in sporadic PD-patient brain samples have yielded conflicting results and the effect of LRRK2 G2019S mutation in human patients has not been assessed (52–55). To determine whether changes in ATP13A2 protein levels were seen in human PD patients that carried the LRRK2 G2019S mutation, we performed western blot analysis of prefrontal cortex samples from five control, five sporadic PD and five LRRK2 G2019S mutation carriers. ATP13A2 levels were increased specifically in LRRK2 G2019S patient samples (Fig. 7C), supporting the link between LRRK2 and ATP13A2 and suggesting that alterations in ATP13A2 may be a LRRK2-mutation specific effect.
LRRK2 regulates the levels of the lysosomal ATPase, ATP13A2, in primary mouse astrocytes, mouse midbrain and human post-mortem samples. (A) ATP13A2 protein levels were compared across Nontg, Wt-LRRK2 Tg, LRRK2 G2019S Tg and LRRK2 KO primary mouse astrocytes. Shown are representative immunoblots with actin as a loading control. Fluorescence signal of immunoblots from multiple experiments was measured and ATP13A2 signals were normalized to actin and expressed as a % of that seen in non-transgenic astrocytes; n ≥ 3. (B) ATP13A2 protein levels were measured from mouse midbrain samples across genotypes, actin normalized and expressed as a % of that detected in non-transgenic midbrain samples; n = 3 animals per genotype. (C) ATP13A2 levels were measured in age-matched, human post-mortem samples from the prefontal cortex of control, LRRK2 G2019S carriers and idiopathic PD patients. ATP13A2 signals were normalized to actin and expressed as a % of that detected in control samples; n = 5 patients per condition. Error bars correspond to SEM and P-values: One-way ANOVA, Bonferroni multiple comparison test; *P < 0.05; **P < 0.01.
Discussion
The present results demonstrate that PD-associated mutations within LRRK2 lead to lysosomal dysfunction and establish that these effects are mediated through LRRK2’s kinase activity. We show that G2019S, the most common mutation associated with LRRK2, results in enlarged lysosomes and reduces the lysosomal capacity of the cell. We establish that this is a common phenotype seen with other PD-associated mutations within LRRK2 and that these lysosomal defects can be reversed by genetic elimination of kinase activity (D1994A co-mutation) and/or LRRK2 autophosphorylation (S1292A co-mutation), suggesting that LRRK2’s kinase activity modulates lysosome morphology through phosphorylation of LRRK2 itself. Further, these findings provide the first evidence, to our knowledge, that treatment with a potent, selective LRRK2 kinase inhibitor can correct defects in lysosome function associated with LRRK2 mutations, highlighting the therapeutic promise of LRRK2 kinase inhibitors in the treatment of PD.
These studies explored the lysosomal role of LRRK2 using primary mouse astrocytes as the main model system, due to their high levels of endogenous LRRK2 protein compared with other cell types assessed and, more practically, because of their amenability for quantitative live-cell imaging experiments. Previous studies have also observed pronounced LRRK2 expression in both primary astrocyte cultures and in astrocytes from human post-mortem brain samples (32,56–58), including recent transcriptome analysis of neurons, glia and vascular cells of the mouse cortex which showed a ≥6-fold enrichment of LRRK2 transcript in astrocytes compared with other cell types examined [(35); http://web.stanford.edu/group/barres_lab/cgi-bin/igv_cgi_2.py?lname=lrrk2], supporting the relevance of astrocytes in the study of physiological and pathological functions of LRRK2. Our work is the first to assess the role of LRRK2 mutations in astrocytes, and additional studies are required to determine whether similar changes in lysosomal morphology are present in patient astrocytes with PD-associated LRRK2 mutations. Further, the potential consequences of astrocytic lysosomal dysfunction on dopaminergic neuron health and survival remain to be explored, and the interplay between astrocytic defects caused by LRRK2 mutation and neuronal homeostasis warrants additional study. Impaired astrocyte function is increasingly recognized as a contributor to neuronal dysfunction and pathology associated with neurodegenerative diseases (59–61), and correcting deficits that affect the ability of astrocytes to promote neuronal survival may represent a valuable therapeutic approach in the treatment of these disorders.
A role for LRRK2 as a regulator of the endolysosomal system has been increasingly appreciated, but a detailed understanding of its precise function in this pathway and whether it promotes or antagonizes lysosomal function under normal (or pathological) conditions was unclear. Previous work has largely focused on the role of LRRK2 in autophagic degradation and has yet to yield a cohesive understanding of its function, with LRRK2 pathogenic mutations or deletion reported to both enhance and impair autophagy depending on the study and model assessed (28,32,36–42). As such, we sought to take a more holistic approach by focusing on the role of LRRK2 on the lysosome itself, as it serves as the endpoint for several degradation mechanisms including the endocytic and autophagic pathways. In the present work, we observed enlarged lysosomal structures in the presence of LRRK2 G2019S mutation, a phenotype which appears to be conserved across other cell types (29,30,62) and is suggestive of defects in lysosome homeostasis. Our functional characterization demonstrated that LRRK2 G2019S mutation does indeed impair lysosome function and leads to a reduction in the lysosomal capacity of the cell. Further, we established that defects in lysosome morphology and function were kinase-dependent and could be reversed both genetically (with kinase dead co-mutation) and pharmacologically (upon treatment with a selective LRRK2 kinase inhibitor). In support of this, cells completely lacking LRRK2 showed an expansion of their lysosomal capacity, with an increase in the average number of lysosomes per cell. It is interesting to note that inhibiting the kinase activity of LRRK2 in non-trangenic and Wt-LRRK2 Tg cells also increased lysosome function, suggesting LRRK2 kinase inhibition may be a more general mechanism in which to improve lysosomal function, even in a non-pathological setting. Together, these data suggest that LRRK2 normally functions as a negative regulator of the lysosomal system and LRRK2 G2019S mutation exacerbates this, leading to lysosomal impairment. While the current work focuses largely on the function of LRRK2 in lysosomal degradation, previous studies have also reported a role for LRRK2 in other protein degradation pathways, including regulating flux through the ubiquitin-proteasome system (63,64). As the interplay between the proteasomal and lysosomal systems has been increasingly appreciated, future work is required to more precisely understand the effect of LRRK2 (and its pathogenic mutations) on the delicate balance of protein homeostasis coordinated by these different degradative pathways.
Perturbations in lysosome homeostasis appear to be a more common phenotype associated with pathogenic mutations within LRRK2, as both mutations in the ROC domain (R1441C) and COR domain (Y1699C) of LRRK2 also lead to lysosomal enlargements. R1441C and Y1699C mutations reduce LRRK2’s GTPase activity and have been reported to elevate its kinase activity (65–70). Our results demonstrated that the introduction of a second (artificial) mutation that ablates kinase activity rescued defects in lysosome size, supporting the hypothesis that pathological increases in LRRK2 kinase activity lead to defects in lysosome morphology and function.
These studies also raise the question of how LRRK2’s kinase activity modulates the lysosomal pathway. While its in vivo substrates remain poorly defined, autophosphorylation of LRRK2 at Ser1292 correlated well with LRRK2’s kinase activity and was increased by PD-associated LRRK2 mutations [the present study and (50,51)]. We show that Ser1292 phosphorylation contributes to the alterations in lysosome morphology seen with LRRK2 mutations, suggesting that LRRK2’s kinase activity is at least, in part, generating its deleterious effects by phosphorylation at this residue. As this autophosphorylation site was only recently identified, many questions remain concerning the downstream functional consequences of Ser1292 autophosphorylation, including whether phosphorylation at this residue can dynamically regulate LRRK2’s subcellular localization and membrane recruitment, ability to dimerize or interaction with other binding partners. Additional LRRK2 substrates or autophosphorylation sites also likely contribute to defective lysosomal morphology as preventing Ser1292 phosphorylation did not completely rescue these defects (compared with kinase-dead mutation). Proteins that regulate lysosome biogenesis, fusion/fission and trafficking represent likely candidates, including reported LRRK2 interactors such as Rab5, Rab7 and components of the cytoskeletal machinery (26,70–75), and potential roles for these proteins in contributing to LRRK2-mediated lysosomal defects warrant further study.
Impaired lysosomal function is increasingly recognized as a unifying feature of many neurodegenerative disorders, and a greater understanding of how lysosomal function is perturbed and potential mechanisms to correct this dysfunction hold great therapeutic promise. Several genes associated with familial forms of PD and inherited parkinsonism mechanistically converge on the endolysosomal system, including the lysosomal cation pump ATP13A2 (8,17). In our search for mechanistic insight into how LRRK2 G2019S mutation lead to a reduction in lysosomal pH, we found that ATP13A2 protein levels were elevated in the presence of LRRK2 G2019S mutation in a variety of systems, including brain tissue from LRRK2 G2019S carriers. These results provide the first evidence for a mechanistic link between LRRK2 and ATP13A2. While this connection between two PD-associated genes is intriguing, many questions remain surrounding the normal function of ATP13A2 and the relationship between these proteins, including whether upregulation of ATP13A2 (and lowering of lysosomal pH seen with LRRK2 G2019S mutation) promotes lysosomal dysfunction or instead acts as a compensatory mechanism to normalize other deficits in lysosome biogenesis or homeostasis. In support of the latter, loss of function mutations in ATP13A2 associated with Kufor–Rakeb syndrome, an early-onset Parkinsonism, have been reported to perturb lysosome function, an effect which was rescued by ATP13A2 overexpression (52,76,77). While determining whether particular lysosomal alterations are detrimental in their own right or are a result of compensatory processes invoked by the cell remains challenging, these studies highlight the adaptive capabilities of the lysosomal system in the face of stress or damage and the therapeutic potential of mechanisms that enhance lysosome biogenesis or activity. Supporting this hypothesis, induction of lysosomal biogenesis by TFEB expression has been shown to attenuate cell death in cellular models of PD (14,78). Approaches that enhance the degradative capacity of the lysosomal system, such as that employed in the present work via inhibition of LRRK2 kinase activity, represent an attractive therapeutic avenue to prevent the protein accumulation seen in PD.
Materials and Methods
Primary astrocyte and neuronal cultures and other cell culture reagents
Previously generated mice expressing the BAC FLAG-Lrrk2-G2019S and FLAG-Lrrk2-Wt transgene in the C57BL/6J background were used in these studies (79). These mice overexpress mouse LRRK2 protein at ∼6- to 8-fold greater level than endogenous mouse LRRK2, and its expression is directed by the endogenous promoter/enhancer regions on the BAC transgene. Additionally, LRRK2 KO mice bred in the C57BL6/J mice background were also used in these studies and have been previously described (80). Astrocyte cultures were generated from cortices of postnatal mice (P1–P3) from non-transgenic, Wt-LRRK2 Tg, LRRK2 G2019S Tg or LRRK2 KO animals by enzymatic and mechanical dissociation. Cell suspensions were plated onto non-coated tissue culture flasks (T150 flasks, BD Biosciences) at a density of one brain per flask and maintained in culture at 37°C in a humidified incubator with 5% CO2. Astrocytes were grown in DMEM (GIBCO) supplemented with 10% heat-inactivated FBS (Atlanta Biologicals, cat #S11550H), 2 mm GlutaMax (GIBCO) and 100 U/mL penicillin and streptomycin (GIBCO). After 1 day in culture, the conditioned media was removed and replaced with fresh growth media. Every 3–4 days after this, half of the media was replaced with fresh growth media. Primary astrocytes were separated from microglia at DIV14 by mechanical shaking (Orbital shaker, 190 rpm, 37°C, 6 h), and the media containing microglia were removed. Astrocytes were subsequently trypsinized using 0.25% Trypsin EDTA (GIBCO) and frozen for subsequent experiments in cell preservation media (Life Technologies).
Primary mouse cortical neurons were prepared from E16 non-transgenic mice. Cortices from embryos were microdissected, trypsinized and dissociated by mechanical and enzymatic methods. Dissociated cortical neurons were counted and plated onto tissue culture treated 6-well plates (BD Biosciences) and maintained in Neurobasal media (GIBCO) supplemented with B27 supplement (GIBCO), 2 mm GlutaMax (GIBCO) and 100 U/mL of penicillin and streptomycin (GIBCO). HEK 293 cells (ATCC) were maintained in DMEM + 10% FBS and 100 U/mL penicillin and streptomycin (GIBCO). All procedures performed on animals in this study were in accordance with regulations and established guidelines, and were reviewed and approved by the Pfizer Institutional Animal Care and Use Committee. Pfizer animal care facilities that supported this work are fully accredited by AAALAC International.
cDNA constructs and transfection
myc-tagged LRRK2, LRRK2 G2019S, LRRK2 R1441C, LRRK2 Y1699C, LRRK2 S1292A and LRRK2 D1994A were synthesized and ligated into pcDNA 3.1 (Gene Dynamics). EGFP-tagged LRRK2 was generated by PCR to amplify the LRRK2 coding sequence and ligated in-frame into EGFP-C1 vector (Clontech). To generate GFP-tagged LRRK2 mutants, mutations were introduced into EGFP-tagged LRRK2 by site-directed mutagenesis using the QuikChange II XL-site-directed mutagenesis kit (Stratagene). All cDNA constructs were verified by sequencing (Beckman Coulter Genomics). LRRK2 mutants were transfected into HEK 293 cells using Lipofectamine 2000 (Life Technologies) or into primary mouse astrocytes using Lipofectamine LTX with PLUS reagent (Life Technologies) and imaged 48 h after transfection. siRNA oligos were obtained from Qiagen (mm ATP13A2 6, 7 and 8) and transfected into ∼50% confluent NIH 3T3 cells using Lipofectamine RNAiMax (Life Technologies).
Spinning-disk confocal microscopy
For live-cell imaging of primary astrocytes, cells were plated onto collagen-coated dishes (Corning and MatTek) 2 days prior to imaging in DMEM without phenol red supplemented with 10% FBS and penicillin and streptomycin. HEK 293 cells were plated onto poly-L lysine coated dishes (MatTek) 1 day prior to imaging. Live-cell imaging was performed using a Zeiss Axio Observer Z1 inverted microscope with a Yokogawa CSU-X1 M 5000 dual camera spinning disk system and acquired with a Photometrics Evolve 512 Delta camera using Zen Blue 2012 software. Cells were visualized using a 60 × 1.4 numerical aperture (NA) objective and illuminated with 405, 445 and 561 laser lines. For fixed-cell imaging, cells were plated onto collagen-coated, 12 mm #1.5 thickness coverslips and later fixed with 4% formaldehyde (Electron Microscopy Sciences) and permeabilized with 0.1% Triton-X-100 in a 1:1 solution of PBS and blocking buffer (Rockland) with 1% normal goat serum (Life Technologies). Cells were labeled using rat anti-Lamp2 (GL2A7, Abcam), rabbit anti-EEA1 (#2411, Cell Signaling), mouse anti-LRRK2 (N241A, UC Davis/NIH NeuromAb Facility), mouse anti-FLAG (clone M2, Sigma) and DAPI (Life Technologies), and subsequently followed by secondary detection using Alexa488-conjugated goat anti-rabbit and Alexa 546-conjugated goat anti-rat (Life Technologies).
Image analysis
Image processing and analysis were performed using ImageJ software (Wayne Rasband, NIH). To measure the size and number of lysosomes, a macro was written in ImageJ that performed background subtraction and thresholding of images and identified lysosomes as individual particles whose signal is above background. These particles were counted and their size measured, and these values are reported on a per cell basis in the figures. Lysosomal size measurements were also binned to score all cells that had lysosomes with a diameter >3 μm to describe the enlarged lysosome phenotype seen upon different LRRK2 manipulations, and these values are described as a percentage of cells within individual experiments that contained any lysosomes with a diameter >3 μm. Lysosome positioning was quantified using a macro written in ImageJ that performed background subtraction and thresholding of images and identified lysosomes as individual particles whose signal is above background. The center of the nucleus was quantified automatically, and the location of these lysosomes with respect to the center of the nucleus was measured. The number of intersections (lysosome signal) for each distance from the nucleus was quantified. To quantify percentage of colocalization, images were acquired sequentially, background subtracted and analyzed using the Coloc 2 plugin in ImageJ.
Post-mortem human samples
Human prefrontal cortex samples were obtained from the University of Miami Brain Endowment Bank. This set consisted of samples from 5 control, 5 sporadic PD and 5 LRRK2 G2019S PD patient samples with an average age of 74.2 ± 4.0, 74.8 ± 3.9 and 74.6 ± 4.0 years, respectively. Additional information concerning these samples can be found in Supplementary Material, Table S1. Samples were homogenized in lysis buffer (50 mm Tris–HCl, pH 7.4, 150 mm NaCl, 1.5 mm MgCl2, 5% glycerol, 1 mm Na3VO4, 25 mm NaF, 1 mm DTT, 0.8% NP-40, 0.5 mm PMSF, supplemented with protease inhibitors), incubated on ice for 30 min and centrifuged at 4°C for 10 min at 20 800 × g. Supernatants were used for subsequent western blot analysis.
Western blot analysis
For cell culture experiments, cells were washed three times with cold PBS and lysed in RIPA buffer (Teknova) supplemented with a standard protease inhibitor cocktail (Roche). Cell extracts were left on ice for 30 min, clarified by centrifugation (20 800 × g for 10 min) and protein concentrations were determined using DC Protein Assay (Bio-Rad). Protein concentrations were normalized for equal loading, and samples were mixed with LDS sample buffer containing sample reducing agent (Life Technologies) to denature samples. Proteins were resolved by SDS–PAGE using 4–12% Bis–Tris gels (Life Technologies) and transferred to nitrocellulose membranes. Membranes were incubated with primary antibodies overnight at 4°C and secondary antibodies (goat anti-rabbit, goat anti-mouse and goat anti-rat) were used at 1:10 000 (LI-COR). Primary antibodies used were anti-LAMP1 (C54H11, Cell Signaling), anti-LAMP2 (GL2A7, Abcam), anti-ATP13A2 (NB110-41486, Novus) and anti-beta actin (AC-74, Sigma). Protein bands were visualized and quantified using the Odyssey infrared scanning system (LI-COR).
Western blot analysis of total and pSer1292 LRRK2
Cells were washed three times with cold PBS and lysed in extraction buffer containing: 50 mm Tris–HCl pH 7.4, 150 mm NaCl, 1.5 mm MgCl2, 5% Glycerol, 0.8% NP-40, 0.5 mm PMSF, PhosStop phosphatase inhibitor cocktail (Roche) and protease inhibitor cocktail (Roche) (81). Cell extracts were left on ice for 30 min, clarified by centrifugation (20 800 × g for 10 min) and protein concentrations were determined using DC Protein Assay (Bio-Rad). Protein concentrations were normalized for equal protein loading, and incubated with LDS sample buffer containing sample reducing agent (Life Technologies) for 10 min at 70°C to denature samples. Proteins were resolved by SDS-PAGE using 3–8% Tris-acetate gels (Life Technologies) and transferred to nitrocellulose membranes. Membranes were incubated with primary antibodies overnight at 4°C and secondary antibodies (goat anti-rabbit or goat anti-mouse) were used at 1:10 000 (LI-COR). Primary antibodies used were anti-LRRK2 (N241A/34, UC Davis/NIH NeuroMab Facility), a previously-described custom antibody against pSer1292 LRRK2 (49) and anti-beta actin (AC-74, Sigma). Protein bands were visualized and quantified using the Odyssey infrared scanning system (LI-COR).
Measurement of lysosomal pH
Lysosomal pH was measured as previously described using the LysoSensor Yellow/Blue DND-160 (Life Technologies) (Coffey, Beckel, Laties and Mitchell (43,82). Briefly, primary astrocytes were plated on collagen-coated, black 96-well plates and labeled with 2 µM LysoSensor Yellow/Blue DND-160 in isotonic solution [Hanks Buffered Salt Solution (GIBCO, Life Technologies) supplemented with 60 mm mannitol]. After 20 min, cells were rinsed three times with isotonic solution and incubated for 10 min with isotonic solution or with pH calibration solutions. pH calibration solutions were generated in 20 mm MES (4-Morpholineethanesulfonic acid), 110 mm KCl and 20 mm NaCl, supplemented with 30 µM nigericin and 15 µM monensin, sterile filtered and titrated to pH values of 4.0, 4.5, 5, 5.5 and 6. The ratio of light excited at 340 nm over 380 nm, detected at a 527 nm emission, was measured at 37°C using a Spectramax M5 plate reader (Molecular Devices), collected every min for 11 cycles and averaged. pH values were then calculated using a standard curve generated from the pH calibrations, plated in adjacent wells and measured simultaneously.
Measurement of cathepsin B activity
To assess cathepsin B activity, a Magic Red fluorescence-based activity assay (ImmunoChemistry Technologies) was used as previously described (83,84). Primary astrocytes were plated on collagen-coated, black 96-well plates with GIBCO DMEM supplemented with 10% FBS and GlutaMax but lacking phenol Red to decrease background signal. Magic Red substrate was reconstituted in DMSO to generate a 150× stock solution and then diluted 1:10 with H2O the day of the experiment. The Magic Red staining solution was added to the cell media to make a 1× solution and incubated at 37°C for 30 min, protected from light. The fluorescence intensity of the cleaved substrate was measured using a Spectramax M5 plate reader (Molecular Devices) with an excitation wavelength of 592 nm and an emission wavelength of 628 nm, using a cut-off filter of 630 nm.
Long-lived protein degradation assay
To assess changes in the degradation of long-lived proteins, a previously established radioactive pulse-chase assay was used (85). Briefly, primary astrocytes were plated 1 × 106 cells per well of a six-well plate and allowed to grow until ∼80% confluent. For LRRK2 inhibitor studies, vehicle or PF-06447475 (500 nM) was pre-incubated for 24 h. L-[2,3,4-3H] valine was added to Neurobasal media supplemented with GlutaMax and penicillin and streptomycin but lacking its own valine (Life Technologies) to a final concentration of 16.7 µM, and the tritiated Neurobasal media was added to cells (with or without PF-06447475 depending on the experiment). Twenty-four hours later, cells were washed three times with Neurobasal media supplemented with 16.7 µM cold valine and GlutaMax and antibiotics. Cells were then incubated in Neurobasal media with cold valine, GlutaMax and antibiotics alone, or with media supplemented with vehicle or 200 µM chloroquine (Sigma) to assess the lysosomal component of long-lived degradation. Vehicle or PF-06447475 was also included for inhibitor studies. After a 24-h chase period, the media was removed and transferred to ultra-clear, pre-lubricated Eppendorf tubes (Corning) on ice. To the media, 100 µl of fresh 10% BSA (Sigma) in PBS followed by 200 µl of room temperature 100% trichloroacetic acid (Sigma) was added to each tube and samples were incubated on ice for 10 min to precipitate proteins. Samples were then centrifuged at 5500 × g for 10 min at room temperature, and the supernatant was collected in glass 20 ml scintillation vials with foil liners (Wheaton Science Products). The supernatant contained the soluble radioactivity which was released from the cells. To the cells remaining in the wells, 1 ml of lysis buffer [10 mm Tris–HCl (pH 8.0), 150 mm NaCl, 2 mm EDTA, 0.5% deoxycholate and 2% NP-40 in H2O] was added, and the cells were placed on a shaker for 10 min at room temperature to aid in cell lysis. The samples were pipetted up and down and transferred to scintillation vials. These samples constitute the total radioactivity contained within the cell. 5 ml of Optiphase HiSafe 3 scintillation cocktail (Perkin Elmer) was added to all samples, and samples were read the next day on a liquid Tri-Carb scintillation counter (Tri-Carb 3100 TR, Perkin Elmer) using 1 min read times. The percentage of proteolysis was determined by taking the ratio of radioactivity released in the media to that contained within the cells.
Statistical analysis
All quantitative data were averaged across multiple experiments, with the number of individual experiments specified in the corresponding figure legend. Error bars represent the standard error of the mean, calculated after compiling mean values across different experiments. Statistical significance was determined using the appropriate variations of one- or two-way analysis of variance (Anova) or with Student's t-test using GraphPad Prism 6 Software. The relative significance for each of the reported differences is specified as P values and are listed in the figure legends and represented graphically within the figures.
Acknowledgements
We would like to thank Kelly Bales, David Riddell, Matt LaVoie and Christine Bulawa for useful comments. We would also like to thank Katherine Hales and Stephen Amato for their technical expertise and Deborah Mash from the University of Miami Brain Endowment Bank for her help with the human post-mortem samples.
Conflict of Interest statement. All authors were employees of Pfizer during the period when the data were generated and interpreted for these studies.
References
Author notes
Present address: Proteostasis Therapeutics, Cambridge, MA 02139, USA.
Present address: Pfizer Vaccines Research East and Early Development, Pearl River, NY 10965, USA.
